Elon Musk’s Case for AI Computing in Space

After the OpenAI verdict, Musk predicts rapid AI growth and lays out the enormous transport needs of space-based computing.

EMElon MuskGast
Kanal
Forbes
Veröffentlicht
19. Mai 2026
Video
26:00
Lesezeit
4 Min.
Original
Englisch
Auf YouTube öffnen
Beim Scrollen schrumpft das Video in die rechte untere Ecke und läuft weiter. Untertitel folgen Ihrer Lesesprache.
LesespracheIn 2 Sprachen verfügbar
Diesen Artikel gibt es noch nicht auf Deutsch, daher wird das Original (Englisch) angezeigt.
ZUSAMMENFASSUNG

In 30 Sekunden

Elon Musk argues that solar-powered AI satellites could eventually draw on far more energy than computing on Earth, but his illustrative plan requires launching 10 million tons to orbit each year. In a May 2026 Forbes interview, he also predicts AI could exceed human abilities in every area within one or two years. Following a court ruling that his OpenAI claims were filed too late, he disputes the finding and says he plans to appeal.

  1. 1Musk contests the timing finding and plans an appealThe jury found his OpenAI claims were filed too late; Musk argues the transformation he challenges took place gradually.
  2. 2A self-growing settlement must survive without Earth shipmentsMusk estimates roughly a million tons might be needed for an industrial base on the Moon or Mars, and ties the plan to reusable rockets.
  3. 3Musk expects AI to exceed human abilities within yearsIn May 2026 he predicts that AI could outperform people in every area within one or two years and expresses concern alongside excitement.
  4. 4Space computing depends on moving enormous massMusk’s illustrative one-terawatt-per-year scenario assumes 10 million tons of annual orbital payload and 100 kilowatts of solar power per ton.

Das Wichtigste

  1. Musk’s illustrative space-computing plan assumes 10 million tons of annual orbital launches for one terawatt of added solar capacity per year.
  2. In May 2026, he predicted AI could surpass humans across all areas within one or two years.
  3. He says a settlement on the Moon or Mars should be able to keep growing if supplies from Earth stop.
  4. Musk disputes the finding that his OpenAI claims were filed too late and says he will appeal.
VOLLTEXT

Den bearbeiteten Text nach Kapiteln lesen

Nur das Transkript
01

The OpenAI verdict and an intended appeal

Musk disputes the finding that he sued too late and argues that the nonprofit-to-for-profit transition has implications for charity.

Forbes interviewer Randall Lane introduces Elon Musk at a gathering of innovators and says Forbes ranked him first on its Innovators 250 list. Lane begins by asking about the verdict in Musk’s case against OpenAI.

Musk says the jury found his claims were brought too late, rather than deciding the underlying questions of unjust enrichment and the nonprofit’s transformation. He disputes that outcome and says the changes happened gradually, making the timing of a possible legal claim difficult to determine.

He argues that OpenAI began with a capped-profit arrangement and an eventual reversion of equity to the charity upon artificial general intelligence, and says the arrangement has changed. Musk calls the conversion a dangerous precedent for charitable giving and says he intends to appeal.

02

Building a civilization beyond Earth

He describes his interests in energy and brain interfaces, then sets out his test for a self-sustaining off-world settlement.

Asked how he came to run several companies, Musk says it mostly happened rather than following a master plan. In college, he wanted to work on things that would change the world, particularly sustainable energy, electric cars, solar power and batteries.

He also describes potential uses for brain-computer interfaces: helping people with brain or spinal injuries communicate, and potentially restoring sight or movement. He presents the broader restoration of vision and walking as possibilities he believes the technology could eventually enable.

Musk hopes SpaceX will achieve a fully reusable orbital rocket later in 2026. In his view, reusable transport is necessary to move enough equipment and material to the Moon or Mars to establish a settlement capable of growing without continued shipments from Earth.

He estimates, as a rough order of magnitude, that such an industrial base might require a million tons. His test for a self-growing settlement is whether it can keep expanding if resupply stops. He adds that becoming multiplanetary means retaining civilization on Earth, not abandoning it for another planet.

03

Choosing work and historical influences

Musk discusses managing several ventures, declines to comment on corporate transactions and names innovators he admires.

When Lane asks how he chooses what to work on each morning, Musk says he does not start the day by deciding what to innovate. He points to ongoing work on space transport, the Starlink satellite network, Tesla’s Optimus robot, self-driving cars and solar power.

Lane asks whether combining the companies might make them easier to coordinate and asks about reports of a SpaceX public offering. Musk declines to discuss either question, citing the constraints around publicly traded companies and a possible quiet period.

Asked about role models, Musk names Nikola Tesla, Thomas Edison and other scientists and inventors. He praises Henry Ford’s contribution to manufacturing complex products at scale and later names Nvidia’s Jensen Huang for work on AI computing.

04

AI forecasts and the case for space power

He predicts rapid AI progress and sketches a solar-powered computing system in space with very large launch requirements.

Musk says he is both excited and frightened by AI’s pace. In May 2026, he predicts AI could exceed human abilities across every area, including innovation, within one or two years; he also voices concern about how such systems will treat people.

Looking toward 2031, he predicts that digital intelligence could surpass the combined intelligence of humanity. He also suggests there might be at least 100 million humanoid robots, perhaps a billion, and that economic output could double over roughly five to seven years.

Lane asks about space data centers and settlements on the Moon and Mars. Musk says SpaceX already has thousands of satellites in orbit and forecasts that Starship could launch more than 10,000 more capable communications satellites a year.

For Musk, the larger opportunity is solar-powered AI computing in space. He argues that the Sun’s energy output is vastly larger than humanity’s present energy use, and that space would allow access to power on a much greater scale.

He gives an illustrative estimate: launching 10 million tons to orbit per year, at 100 kilowatts per ton, could add a terawatt of solar-powered AI capacity annually. He compares that power level with average US electricity use, framing it as a physical possibility rather than an existing system.

For still greater scale, Musk proposes making solar panels and radiators on the Moon and launching them with a mass driver, a device that accelerates payloads without a conventional rocket. He suggests chips might initially come from Earth, and estimates a possible future build-out of a petawatt per year.

05

Tunnels, synthetic medicine and a legacy

Musk identifies underground transport and programmable medicine as opportunities, then describes how he hopes to be remembered.

Asked for ideas he has not yet pursued, Musk returns to tunnels. He says a transport network with multiple underground levels could address traffic associated with dense buildings, and encourages others to build tunneling companies alongside his Boring Company.

He also sees an opportunity in synthetic medicine. Musk compares older drug discovery to finding useful objects by chance and argues that custom-made RNA could make future medicine more programmable. He speculates that, with the right instructions, synthetic RNA might eventually treat a very wide range of conditions.

Musk briefly mentions electric aircraft as another area of opportunity. Asked what he would like his legacy to be, he answers that he hopes to have played a useful role in advancing civilization. Lane thanks him and returns the event to its other honorees.